Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Slug catchers sizing 2
- Thread starter mucour
- Start date
Try the Gulf publishing site. They may have an old article from Hydrocarbon Processing.
Also, remember that some peple may call them knock out pots or disentrainment drums so be generous when searching.
API RP-521 has a section on Knock out which is basically the Stokes equation which you can also find in "Perry" (Chem eng handbook).
If the web references don't go back far enough you'll probably have to go to an engineering library. ASME have a good one in New York (but if you're not in NY that won't help)
Also, remember that some peple may call them knock out pots or disentrainment drums so be generous when searching.
API RP-521 has a section on Knock out which is basically the Stokes equation which you can also find in "Perry" (Chem eng handbook).
If the web references don't go back far enough you'll probably have to go to an engineering library. ASME have a good one in New York (but if you're not in NY that won't help)
- Thread starter
- #3
Flareman, I have gone through API RP-521 and it does not explain to me how to estimate the slug volume expected from the pipeline. This data is required for the vessel sizing for it to be able to handle slug.
I quite agree with you that a slug catcher is similar to any pressure containng vessel like knock-out pot, but the challenge is in determining the slug volume.
I have carried out pipeline simulation in the past using Pipephase and as part of the computer result print-outs it normally give the slug volume (or is it sphere voluume, I can't remember anymore).
Is this the only way of determining the slug volume?
Are there other mathematical methods of estimating this slug volume, based on certain assumptions?
I quite agree with you that a slug catcher is similar to any pressure containng vessel like knock-out pot, but the challenge is in determining the slug volume.
I have carried out pipeline simulation in the past using Pipephase and as part of the computer result print-outs it normally give the slug volume (or is it sphere voluume, I can't remember anymore).
Is this the only way of determining the slug volume?
Are there other mathematical methods of estimating this slug volume, based on certain assumptions?
Predicting/setting the potential slug volume has a large factor of experience in it.
Pipephase does predicts the slug volume, it's related to the probability of that size slug arriving, similar to a 20 year storm, 50 years storm, 100 year storm, etc. It's been a while since I ran Pipephase so I can't be certain if it's something that is always printed out or if you have to select it as an additional output. And I have no knowledge to say how Pipephase's results matches reality.
The sphere volume in Pipephase is essentially how much liquid you would get if you ran a pig or sphere through the pipeline and displaced the liquid hold up in the pipeline out the end.
There was an article recently on a new method to predict and control slug volume in either Chemical Engineering magazine or Chemical Engineering Progress (the AIChE mag). I've also seen control valves used to slow down a arriving slug and decrease the flow rate into the slug catcher to give you time to also be processing the liquids out of the slug catcher as more is coming in. Sizing the catcher for the maximum slug you determine is one thing, but if it's a huge vessel, there is going to be lots of pushing to reduce it 'somehow' and employ other methods to detect and control slugs rather than just sheer volume.
Pipephase does predicts the slug volume, it's related to the probability of that size slug arriving, similar to a 20 year storm, 50 years storm, 100 year storm, etc. It's been a while since I ran Pipephase so I can't be certain if it's something that is always printed out or if you have to select it as an additional output. And I have no knowledge to say how Pipephase's results matches reality.
The sphere volume in Pipephase is essentially how much liquid you would get if you ran a pig or sphere through the pipeline and displaced the liquid hold up in the pipeline out the end.
There was an article recently on a new method to predict and control slug volume in either Chemical Engineering magazine or Chemical Engineering Progress (the AIChE mag). I've also seen control valves used to slow down a arriving slug and decrease the flow rate into the slug catcher to give you time to also be processing the liquids out of the slug catcher as more is coming in. Sizing the catcher for the maximum slug you determine is one thing, but if it's a huge vessel, there is going to be lots of pushing to reduce it 'somehow' and employ other methods to detect and control slugs rather than just sheer volume.
Sorry Mucour but we're moving into an area I can't really help with.
My knowledge of two-phase flow systems suggests that this becomes a somewhat complex interaction between the equilibruim state in the pipeline, the pipeline diameter and length, and flow regimes.
Slug flow is a specific two phase condition at relatively low velocities. My own two phase experience of hold up prediction is in the mist flow regime. I suggest you get a book on two-phase flow (Piping Handbook has some references) or, better still, pray that someone else on the site has a handy solution.
Sorry!![[sad]](/data/assets/smilies/sad.gif "[sad] [sad]")
My knowledge of two-phase flow systems suggests that this becomes a somewhat complex interaction between the equilibruim state in the pipeline, the pipeline diameter and length, and flow regimes.
Slug flow is a specific two phase condition at relatively low velocities. My own two phase experience of hold up prediction is in the mist flow regime. I suggest you get a book on two-phase flow (Piping Handbook has some references) or, better still, pray that someone else on the site has a handy solution.
Sorry!
-
2
- #6
- Thread starter
- #8
fdomin, I am engaged in the same problem about predicting a slug catcher after running HYSYS / PIPESYS software.
Then I will appreciate very much if you can send me the GPSA article you have mentiones above.
MY EMAIL IS:
attiliopraderio@cantv.net
Also if TD2K could remember a more exact address to find out the article, it will be also appreciated.
Finally, MortenA, my case is a 200 kms 16 inch pipeline transporting a retrograde gas coming from an offshore platform and arriving to shore at the entrance of a NGL plant, Any ideas or rules of thumb similar to the vertical riser you explained? Your help to me in the corrosion thread was very helpful .. Thanks.
Then I will appreciate very much if you can send me the GPSA article you have mentiones above.
MY EMAIL IS:
attiliopraderio@cantv.net
Also if TD2K could remember a more exact address to find out the article, it will be also appreciated.
Finally, MortenA, my case is a 200 kms 16 inch pipeline transporting a retrograde gas coming from an offshore platform and arriving to shore at the entrance of a NGL plant, Any ideas or rules of thumb similar to the vertical riser you explained? Your help to me in the corrosion thread was very helpful .. Thanks.
A1J2,
I can suggest 3 approaches to estimate the size of a slug catcher. You may want to take the largest of the 3 and apply an appropriate safety factor:
1) As posted previously, the volume of the statisically largest 1/1000 slug.
2) Volume swept by pig (sphere), also posted previously
3) Cunliffe's method:
Cunliffe’s Method which is useful to predict the liquid surge rate due to an overall gas rate change for condensate pipelines. This method is particularly useful for estimating liquid handling capacity for ramp-up (increasing gas rate) cases. As the gas rate increases, the total liquid holdup in the line will drop owing to less slippage between the gas and liquid phases. The liquid residing in the line is therefore accelerated to the equilibrium velocity at the final gas rate and thus expelled at a rate higher than the final equilibrium liquid rate for the duration of the transition period. The transition period is assumed to be equal to the residence time at the final gas rate, that is, the time it takes the liquid to travel from one end of the line to the other.
The average liquid rate during the transition period can be determined as follows:
QLt = QLi +(HL tot-I – HL tot-F)/ tr
QLi = QGi(LGRout)
tr = HL tot-F/Qci
Where,
QLt = liquid rate during the transition period
QLi = initial condensate rate
QGi = initial gas rate
HL tot-I = total liquid holdup volume in line – initial gas rate
HL tot-F = total liquid holdup volume in line – final gas rate
LGR = liquid/gas ratio at outlet pressure (assumed constant)
tr = liquid residence time (at final flowrate)
Cunliffe tested this method with field measurements for a 67 mi. 20 in. pipeline with an average operating pressure of 1300 psig and an LGR of 65 bbl/MMscf. He found that the change in condensate flow rate can be predicted to within 15% using this method.
Reference: Cunliffe, R.: “Prediction of Condensate Flow Rates in Large Diameter High Pressure Wet Gas Pipelines”, APEA Journal (1978), 171-177.
Methods 1 and 2 are reported as output from a PIPESIM model similar to PIPEPHASE as described in a previous post. Cunliffe's method requires a quick hand calculation. If you are using Hysys, you may want to consider embedding a PIPESIM sumulation model into Hysys as a unit op to assist in sizing your slug catcher. For more info on doing this, see:
Regards,
Mack
I can suggest 3 approaches to estimate the size of a slug catcher. You may want to take the largest of the 3 and apply an appropriate safety factor:
1) As posted previously, the volume of the statisically largest 1/1000 slug.
2) Volume swept by pig (sphere), also posted previously
3) Cunliffe's method:
Cunliffe’s Method which is useful to predict the liquid surge rate due to an overall gas rate change for condensate pipelines. This method is particularly useful for estimating liquid handling capacity for ramp-up (increasing gas rate) cases. As the gas rate increases, the total liquid holdup in the line will drop owing to less slippage between the gas and liquid phases. The liquid residing in the line is therefore accelerated to the equilibrium velocity at the final gas rate and thus expelled at a rate higher than the final equilibrium liquid rate for the duration of the transition period. The transition period is assumed to be equal to the residence time at the final gas rate, that is, the time it takes the liquid to travel from one end of the line to the other.
The average liquid rate during the transition period can be determined as follows:
QLt = QLi +(HL tot-I – HL tot-F)/ tr
QLi = QGi(LGRout)
tr = HL tot-F/Qci
Where,
QLt = liquid rate during the transition period
QLi = initial condensate rate
QGi = initial gas rate
HL tot-I = total liquid holdup volume in line – initial gas rate
HL tot-F = total liquid holdup volume in line – final gas rate
LGR = liquid/gas ratio at outlet pressure (assumed constant)
tr = liquid residence time (at final flowrate)
Cunliffe tested this method with field measurements for a 67 mi. 20 in. pipeline with an average operating pressure of 1300 psig and an LGR of 65 bbl/MMscf. He found that the change in condensate flow rate can be predicted to within 15% using this method.
Reference: Cunliffe, R.: “Prediction of Condensate Flow Rates in Large Diameter High Pressure Wet Gas Pipelines”, APEA Journal (1978), 171-177.
Methods 1 and 2 are reported as output from a PIPESIM model similar to PIPEPHASE as described in a previous post. Cunliffe's method requires a quick hand calculation. If you are using Hysys, you may want to consider embedding a PIPESIM sumulation model into Hysys as a unit op to assist in sizing your slug catcher. For more info on doing this, see:
Regards,
Mack
First, run pipephase to model your application. Forget, hysim as it is not suitable for pipeline hydraulics.
For pipe type slug catchers contact Taylor Forge. They sell a complete piping assembly and can provide some assistance.
Vessel type slug catchers can only be used if the incoming liquid volume is small. Otherwise, the size of the vessel(s) is uneconomical because it will be hugh.
You first need to do some reading on two phase flow patterns. Depending on the holdup factor you may have to run several correlations and then apply a judgement factor. See Beggs and Brill's book available from the Univ. of Tulsa and the pipephase manual for guidance.
As I recall, unless the holdup factor is greater than about 50% then you are not going to have a flow induced slug. So the problem becomes how much liquid do you receive when you run a pig or how much liquid do you receive when you increase the flow rate. Higher flow rates cause a lower holdup factor as noted previously and the difference comes out the end of the pipeline.
After you know how much liquid to design for then you can start to design the slug catcher. Often the design volume involves a "pucker" factor depending on your confidence of the holdup correlation as there is allways someone looking over your shoulder who is concerned with the installation economics.
Don't forget you have to handle the gas in the slug.
Finally, sizing slug catchers involves a lot of experience as well as technology. I have taken two graduate level courses in two phase flow and it only gave me a start.
For pipe type slug catchers contact Taylor Forge. They sell a complete piping assembly and can provide some assistance.
Vessel type slug catchers can only be used if the incoming liquid volume is small. Otherwise, the size of the vessel(s) is uneconomical because it will be hugh.
You first need to do some reading on two phase flow patterns. Depending on the holdup factor you may have to run several correlations and then apply a judgement factor. See Beggs and Brill's book available from the Univ. of Tulsa and the pipephase manual for guidance.
As I recall, unless the holdup factor is greater than about 50% then you are not going to have a flow induced slug. So the problem becomes how much liquid do you receive when you run a pig or how much liquid do you receive when you increase the flow rate. Higher flow rates cause a lower holdup factor as noted previously and the difference comes out the end of the pipeline.
After you know how much liquid to design for then you can start to design the slug catcher. Often the design volume involves a "pucker" factor depending on your confidence of the holdup correlation as there is allways someone looking over your shoulder who is concerned with the installation economics.
Don't forget you have to handle the gas in the slug.
Finally, sizing slug catchers involves a lot of experience as well as technology. I have taken two graduate level courses in two phase flow and it only gave me a start.
MMcLean
Chemical
- Jul 21, 2003
- 25
- 0
- 0
Hi
I'm interested in in getting a copy of the paper by Cunliffe that BusyCEO (Mack) mentioned in his discussion. I have a copy but some of the formulae in Eaton's correlation are not very legible. Could you email it to me if you have a copy please.
Thanks
Mark
I'm interested in in getting a copy of the paper by Cunliffe that BusyCEO (Mack) mentioned in his discussion. I have a copy but some of the formulae in Eaton's correlation are not very legible. Could you email it to me if you have a copy please.
Thanks
Mark
ggordonstewart
Chemical
Gas Conditioning and Processing, Vol.1: The Basic Principles by John M. Campbell, published by: John M. Campbell and Company has a good section on 2-phase flow in both horizontal and vertical pipes.
If you have access to a good process simulator such as Hysys, you can calculate the liquid hold-up under different flow conditions provided you have a pipeline profile (elevation changes, distances and piping size. Pipesys may be better at determining this.
If all else fails, the worst slug is the total liquid produced daily, provided that the line is pigged daily. This is sometimes a smaller slug than that calculated to occur when the flow rate increases.
Remember that these are estimates only. Science is not very exact in this area. So, err on the side of caution.
G. Gordon Stewart, P.Eng.
Gas & Oil Process Engineering Consultant
ggstewar@telusplanet.net
If you have access to a good process simulator such as Hysys, you can calculate the liquid hold-up under different flow conditions provided you have a pipeline profile (elevation changes, distances and piping size. Pipesys may be better at determining this.
If all else fails, the worst slug is the total liquid produced daily, provided that the line is pigged daily. This is sometimes a smaller slug than that calculated to occur when the flow rate increases.
Remember that these are estimates only. Science is not very exact in this area. So, err on the side of caution.
G. Gordon Stewart, P.Eng.
Gas & Oil Process Engineering Consultant
ggstewar@telusplanet.net
Average Slug Length : 350 pipe ID (Ref Scott, Soham & Brill 1986)
Further Reading
SPE15103 : Prediction of Slug Length in Horizontal Large Diameter pipes (Scott, Soham & Brill)
SPE20645 : Gas Liquid Flow in Pipelines
SPE27960 : Slug Flow, Occurrence, Consequences & Prediction
SPE39856 : A study of Slug Characteristics for Two Phase Horizontal Flow
OTC 7744 – Slug Sizing / Slug Volume Prediction, State of the Art Review & Simulation
Further Reading
SPE15103 : Prediction of Slug Length in Horizontal Large Diameter pipes (Scott, Soham & Brill)
SPE20645 : Gas Liquid Flow in Pipelines
SPE27960 : Slug Flow, Occurrence, Consequences & Prediction
SPE39856 : A study of Slug Characteristics for Two Phase Horizontal Flow
OTC 7744 – Slug Sizing / Slug Volume Prediction, State of the Art Review & Simulation
Demethanizer
Chemical
Really thanks to everty body
Regarding all you said about the Pipesys extension and pipephase method for finding the sphere volume (or accumulated liquid hold-up) I should remember you that it will give us the result of slug volume at normal condition and also pigging condition which is useful for an estimation of slug catcher sizing. More research have to be done with dynamic simulation (like OLGA2000) when the pipeline is starting-up after a long shut-down and the fluid temperature inside the pipeline is cooled down to ambient Temp. Slugtracking module in OLGA will provide us slug volume and frequency. These strategy has been carried out in South Pars gas field by FW and scandpower.
I myself have a less knowlede than you guys in two phase flow, I really appreaciate anyone who (ggordonstewart, busyCEO (mack) and a1j2) could help me for selecting the best 2-phase correlation for a gas based simulation in Pipesys which really impacts the result of slug volume. I mean the correllations like Eaton et al. - Beggs and Brill - Dukler - OLGAS - hughmark or... for Liquid hold-up and friction loss.
Thanks
Demethanizer
Regarding all you said about the Pipesys extension and pipephase method for finding the sphere volume (or accumulated liquid hold-up) I should remember you that it will give us the result of slug volume at normal condition and also pigging condition which is useful for an estimation of slug catcher sizing. More research have to be done with dynamic simulation (like OLGA2000) when the pipeline is starting-up after a long shut-down and the fluid temperature inside the pipeline is cooled down to ambient Temp. Slugtracking module in OLGA will provide us slug volume and frequency. These strategy has been carried out in South Pars gas field by FW and scandpower.
I myself have a less knowlede than you guys in two phase flow, I really appreaciate anyone who (ggordonstewart, busyCEO (mack) and a1j2) could help me for selecting the best 2-phase correlation for a gas based simulation in Pipesys which really impacts the result of slug volume. I mean the correllations like Eaton et al. - Beggs and Brill - Dukler - OLGAS - hughmark or... for Liquid hold-up and friction loss.
Thanks
Demethanizer
- Status
Similar threads
- Question
- Question
- Locked
- Question